CHAPTER 5

93

The syn- and post-collisional evolution of theCarpathians foredeep (Romania): newconstraints from AMS and paleostress analyses

AbstractWe present here the first anisotropy of the magnetic susceptibility (AMS) data from the

Romanian Carpathians foredeep, which generally reveal radial compression directions

perpendicular to the Carpathians orogenic arc. Distribution of the present-day stress

field in the Pannonian-Carpathians region basically reflects the pre-imposed plates

boundaries, established during late Miocene collision. Analysis of the AMS along the

eastern and southern Carpathians foredeep, at the contacts of the orogenic nappe pile,

the lower plate, and the overlying sedimentary rocks, indicates that two factors are

critical for the distribution of the stress field during and postdating the collision, i.e. 1)

the inherited highly bended plate geometry and 2) the Quaternary deformations leading

to differential vertical and horizontal movements in the SE Carpathians.

This chapter was co-authored by Iuliana Vasiliev, Liviu Matenco and Wout Krijgsman

94

Figure 1. (a) Tectonic map of the Eastern Alps–Carpathians–Dinarides–Balkans region (afterSchmid et al. 2006). (b) Transect across the Focsani basin illustrating the geological contacts withinthe study area (after Matenco et al. 2006). The route of the section is represented with dashed linein the panel a.

1. IntroductionSubduction and continental collision of highly arcuate orogens is a non-cylindrical

process where deformation is expressed through a complex interplay between (oblique)contraction and strike-slip, abnormal foredeep geometries and contrasting patterns ofvertical movements (Bertotti et al. 2001; Faccenna et al. 2002; Foeken et al. 2003). Duringpost-collisional times, on-going deformation acting on the locked plate boundary respondsto changes in the regional intra-plate deformation (Horvath 1993) due to far-field stressesand can result in significant deformation such as crustal and/or lithospheric folding(Cloetingh & Burov 1999). Significant deformation can result in post-collisional timesfrom memory effects inherited from the oceanic subduction stage via deep mantle processes,such as slab detachment (Wortel & Spakman 2000), delamination (Sacks & Secor 1990),or thermal re-equilibration (Toussaint et al. 2004). The interplay between these deep andshallow processes in highly arcuate settings create complicated deformation patterns at thelocal scale, but is generally the overall result of the same stress field acting at the orogenicscale.

Because the post-tectonic covers are often missing in thin-skinned orogenic areasdue to erosion active during the (post)collisional rebound stage, a discrimination betweenvarious coeval tectonic processes acting at the plate boundaries is not straightforward andmostly relies on isotope geochronology (Dempster & Persano 2006), which is usually oneorder of magnitude less precise than direct stratigraphic markers. These processes can

95

therefore be ideally studied in “soft” collisional orogens (dominated by subduction), wherethe low amount of topography generated during the shortening stage is subsequently burieddue to significant subsidence in the post-collisional stages, i.e. orogens which have the post-tectonic covers still preserved, such as the the Apennines (Bertotti et al. 2001) or theCarpathians (Matenco et al. 2006).

One of the critical questions concerning the highly arcuate shape of the easternand south Carpathians arc is the distribution of the crustal stress field during the latestMiocene – Quaternary times, i.e. after the nappe stacking, related to the active subductionof the distal parts of the lower Moesian/Scythian/Eastern-European plate, has ended inearly-late Miocene (~11Ma).

Recent studies on the geometry of the SE Carpathians foreland (Tarapoanca et al.2004) and its post-collisional evolution and impact on the present-day movements (Leeveret al. 2006; Matenco et al. 2006: Schmid et al. 2006), as detected by modern GPS studies(van der Hoeven et al. 2005), have revealed a consistent SE-ward movement of the blockcomprised between the Intramoesian and ~Trotus faults (Fig. 2). Quaternary shortening inthe order of 5 km (Leever et al. 2006) generated differential vertical movements. Significantuplift occurred of an area juxtaposed roughly on the present day mountain chain (Mertenet al. 2005), being comparable with the subsidence in the foreland. The inferred shorteningdirection is WNW-ESE (Matenco et al. 2006), which has induced pulses of differentialvertical movements migrating in space and time during the Quaternary (Necea et al. 2005).

Several paleostress studies infer that during the Pliocene-Quaternary a N-S orienteddirection of shortening was active for the South Carpathians (Matenco et al. 1997a; Hyppoliteet al. 1999), NW-SE to NNW-SSE compression for the junctions with the eastern Carpathians(Hyppolite & Sandulescu 1996; Morley 1996) and N-S to NNW-SSE shortening for thecentral-northern part of the eastern Carpathians (Matenco & Bertotti 2000). Like alwaysin this paleostress methodology, the timing constraints are rather weak, the main bulk ofthe deformation being measured in pre-Miocene sediments, due to little preservation ofkinematic indicators and/or reduced exposure of the poorly consolidated Miocene-Quaternary sediments. This makes it rather difficult to separate the stress fields activeduring the Carpathians collision from the ones, which largely postdate it.

The anisotropy of the magnetic susceptibility (AMS) has earlier proven to be veryuseful to examine the patterns of the strain in orogenic settings like the Aegean arc,Apennines and Sardinia (Kissel et al. 1986; Sagnotti & Speranza 1993; Scheepers & Langereis1994; Mattei et al. 1997; Duermeijer et al. 1998; Mattei et al. 1999; Duermeijer et al. 2000;Faccenna et al. 2002). Analysis of the AMS can be used to establish the sedimentary andtectonic history in weakly deformed sediments, because of its relationship with the regionalstress field (Tarling & Hrouda 1993). Upon deformation, the lineation given by the maximumaxes of AMS quickly aligns along the direction of maximum extension or, equivalently,perpendicular to maximum compression. Because of the availability in young, weaklydeformed sediments of young age, in which paleostress indicators are usually missing, thistype of indicators represents a powerful tool for extracting stress states, especially whencombined with other kinematic observations from structural studies.In this paper we present the first AMS results from the late Miocene-Pliocene sedimentarysuccessions of the Carpathians foredeep of Romania. We attempt to discriminate at thescale of the entire Romanian Carpathians to distinguish between late Miocene kinematicdirections associated with the last thrust nappe emplacement of the Carpathians from theQuaternary folding episode focused in the SE-most corner and its associated present-day

96

Figure 2. (next page) Simplified tectono-structural map of the Carpathians modified after thegeological maps of 1:200000 of Romania. Compilation of the paleostress measurements per-formed in the eastern and southern Carpathians (Hyppolite & Sandulescu 1996; Morley 1996;Matenco et al. 1997a; Matenco & Schmid 1999; Matenco et al. 1997b; Zweigel et al. 1998). In theeastern Carpathians Foredeep, the contours of –500, –1000 and –1500 m indicate the depth ofthick quaternary deposits in the Focsani Basin (after Matenco et al. 2006).

effects both in the shallow crust and at deep mantle level. In addition, the AMS analysescan be used as time indicators for the recent tectonic evolution, because all results will bederived from well-dated sedimentary rocks (Vasiliev et al. 2004; Vasiliev et al. 2005).

2. Syn- and post-collisional evolution of the eastern and southernCarpathians

The Romanian Carpathians represent an arcuate belt formed in response to theTriassic to Tertiary evolution of three continental blocks (Fig. 1). The first two are referredto as Internal and Median Dacides, respectively (Sandulescu 1980; Sandulescu 1988), alsoreferred to as “Tisza” and “Dacia” blocks (Fig. 1a) (Balla 1986; Csontos & Vörös 2004),Schmid et al. 2006) and are found in the west and south. The third one is formed by theEastern European, Scythian and Moesian platforms found to the eastern and north(Sandulescu 1984; Sandulescu & Visarion 1988; Visarion et al. 1988) (Fig. 1a). Theseblocks were formerly separated by two oceanic domains, whose remainders are found inthe Transylvanides (Mures-Internal Vardar zone) to the west and south, and the OuterDacidian trough (Ceahlau-Severin ocean) to the eastern and north (Sandulescu 1984). Themore external, partial oceanic, basin of the Outer Dacidian basin formed in the lateJurassic and evolved in a passive margin setting in respect to the Median Dacides throughoutthe Early Cretaceous. The only sedimentary cover of this domain still preserved, andknown to have been deposited over an oceanic basement, is of Upper Jurassic – LowerCretaceous age. It is still preserved in the Ceahlau-Severin nappe, which was successivelythrusted during the intra-Albian and intra-Senonian tectonic events (Stefanescu & Group1988; Iancu et al. 2005). As a result, the slab of the Outer Carpathians is ~160-105 Maold and has been completely subducted at ~75 Ma. The younger sediments of this basinstill preserved in the outer flysch belt (Moldavides) were exclusively deposited over thinnedcontinental crust of the eastern passive continental margin with respect to the OuterDacides (Micu 1990). These younger sediments were detached from their original basementby thin-skinned thrusting and transpression during the Miocene orogeny.

2.1. Collision in the eastern and south CarpathiansThe evolution of nappe stacking of the Moldavides in the eastern Carpathians

(Fig. 1) is well documented (Sandulescu 1988; Roure et al. 1993). Following Early andMiddle Miocene nappe stacking events leading to the emplacement of the InternalMoldavides, thrusting culminated in the early late Miocene (late Sarmatian s.l., ~11Ma),when the Subcarpathian nappe was thrust on top of the sedimentary cover of the forelandplatforms (Sandulescu 1988; Matenco & Bertotti 2000). The chain is non-cylindrical, thegeometry and kinematics of thrusting change along strike as a consequence of the pre-existing structural grain (Ellouz & Roca 1994), and the total shortening across the easternCarpathian outer units (Moldavides) in Late Oligocene to Late Miocene amounts to ~160km (Roure et al. 1993; Ellouz & Roca 1994).

97

The Tertiary tectonic evolution of the South Carpathians/Balkans system isdominated by the large scale rotation of the Tisza-Dacia unit around Moesia during thePaleogene-Early Miocene (Csontos & Vörös 2004; Fugenschuh & Schmid 2005) and itssubsequent Middle-Late Miocene indentation against the European margin, presumablydriven by the roll-back, subduction and detachment of the distal parts of the European/Scythian/Moesian margins (Royden 1988; Wortel & Spakman 2000). At the connectionbetween the Carpathians and the Dinarides, the first period of rotation is associated withthe last phase of Paleogene (transpressional) deformation in the Balkans (Doglioni et al.1996), accompanied in the South Carpathians by rotation through orogen parallel extensionduring the latest Cretaceous – Eocene (Schmid et al. 1998; Matenco & Schmid 1999;Fugenschuh & Schmid 2005) and dextral movements along curved faults systems. Recentinterpretations based on seismic lines (Rabagia & Matenco 1999; Tarapoanca 2004) haverevealed that the direct prolongation of the Timok system towards north and eastern doesnot rely on a single discontinuity, but rather to a large scale Early Miocene transtensionalsystem. Middle and late Miocene times represent periods of basin inversion, docking ofthe South Carpathians against the Moesian Platform and the onset of the Subcarpathianthin-skinned nappe unit at their contact (Dicea 1996). While the Middle Miocene (Badenian)is characterised mostly by internal shortening and the apparent thrusting of the Danubian/Getic system over the sediments of the Getic Depression (Matenco et al. 1997b; Stefanescu& Group 1988), the late Miocene represents the main moment when the Subcarpathiannappe was thrusted on top of the Moesian platform (Sandulescu 1988), being associatedwith an overall dextral translation of the internal South Carpathians with respect to Moesia.In the Getic Depression, the late Miocene deformation is organised in an overall transferof significant dextral strike-slip deformations in the western part towards a graduallyincreased amount of thrusting in the eastern (Matenco & Schmid 1999), reaching ~40 kmoffset near the Intramoesian fault (Fig. 1). Conjugate sinistral faulting slightly postdates theonset of the main thrusting and relates to ongoing strike-slip deformation after the mainplate boundary has been locked down (Rabagia & Matenco 1999). Paleomagnetic resultsindicate that systematic ~30° clockwise rotations occurred in the southern Carpathiansafter ~13 Ma, and that tectonic rotation has generally ceased after ~9 Ma (Dupont-Nivetet al. 2005).

One common feature characterizes both the eastern and South Carpathians(oblique) shortening evolution. The main collisional phase, i.e. the moment when the non-thinned lithospheric part of the lower plate arrived at the subduction zone, is common inboth areas and is of late Miocene age (late Sarmatian s.l.). The orogenic system over-thickened and subduction/underthrusting stopped. This is the moment when the backarcbasins (Pannonian) opened as a result of the overall E-ward movement of the Carpathiansupper plate (Schmid et al. 1998), driven by the subduction and roll-back of the lower platewhich inverted (Horvath 1993). It is generally accepted that this collision moment iscoeval with the arrival at the subduction zone of the thick, cold and buoyant Eastern-European craton in the central-northern part of the eastern Carpathians (Horváth &Cloetingh 1996).

2.2. Post-collisional evolutionLarge-scale deformation is also recorded during the latest Miocene-Quaternary

post-collisional evolution, in particular in the SE Carpathians and the adjacent Focsanibasin. The spectacular 13 km deep Miocene Focsani basin (Fig. 1b) (Tarapoanca 2003)

98

shows a thick post-orogenic infill on top of the Late Miocene syn-tectonic sediments.Moreover, the strata steeply dip away towards the foreland at the western flank of thisbasin (Fig. 1b), which is in contradiction with the standard foredeep geometry (Leever et al.2006).

Typically, along the entire sector of the eastern and southern Carpathians wherethe Moesian platform represents the foreland, the frontal part of the thin-skinned nappepile is covered by post-collisional uppermost Miocene to Quaternary deposits with up to 5km thickness. The particularly large subsidence in the Focsani basin, associated with largescale tilting on its western flank, results from a Quaternary-age interference of a crustalfolding mechanism acting only in a restricted sector of the chain, between the Intramoesianand Trotus faults (Fig. 2) (Cloetingh et al. 2004). As a result two individual deformationepisodes are observed during the post-collisional evolution of the Carpathians, latestMiocene – Pliocene subsidence and Quaternary folding. These represent an effect of theinterplay between two mechanisms, the pull-down effect of a slab, inherited and lockedduring the late Miocene Carpathians collision, and the Quaternary inversion of the entireCarpathians-Pannonian system (Bada et al. 1999; Pinter et al. 2005).

A large number of earthquakes cluster in the SE Carpathians, mostly focused inthe so-called Vrancea area. The volume of 80 x 40 x 210 km exhibiting intermediatemantle seismicity has recently been the subject of numerous studies (Oncescu & Bonjer1997; Wenzel et al. 1998). The intermediate mantle seismicity is commonly interpreted interms of slab-pull exerted by subducted oceanic lithosphere forming high-velocity body(ies)identified by regional seismic tomography studies (Wortel & Spakman 2000). In contrastwith the intermediate-mantle seismicity, the mechanisms of the also significant crustalseismicity are different, underlying the large scale Pliocene-Quaternary folding observedin the upper crust (Matenco et al. 2006). Focal mechanism solutions do not provide aconsistent crustal stress regime: they are dispersed and mostly indicate compression andstrike-slip on the western Focsani flank, respectively extension on the eastern flank. Theseobservations are compatible with the GPS measurements acquired in the RomanianCarpathians (van der Hoeven et al. 2005), which suggest consistent horizontal and verticaldisplacements, post-seismic deformation linked to the large Vrancea earthquakes beingapparently minor. A motion of the Moesian block towards ESE is observed, displacementsof ~3 - 4 mm/y being laterally bounded by the Intramoesian and Trotus Faults, compatiblewith the WNW-ESE direction of the Pliocene-Quaternary folding (van der Hoeven et al.2005).

2.3. The syn- and post-collisional evolution of the (paleo)stress field and inferred kinematicdirections

Kinematic studies performed in the eastern Carpathians (Fig. 2) (Hyppolite &Sandulescu 1996; Morley 1996; Zweigel et al. 1998; Matenco & Bertotti 2000) indicatedtwo main deformation stages: one in the Upper Miocene and the other one is Plio-Quaternary. Miocene shortening led to major thrusting and folding of the Moldavidesnappes. The direction of compression changed gradually from WSW-ENE in the centraleastern Carpathians (Matenco & Bertotti 2000), to WNW –ESE in the bending area andsouthwards further to NW-SE (Hyppolite & Sandulescu 1996; Morley 1996; Zweigel et al.1998) (Fig. 2), and is thus parallel with the transport direction. Towards the end of theSarmatian collision, a change in the stress field leading to widespread strike-slip deformationis observed in all areas of eastern Carpathians, ~E-W sinistral north of the Trotus fault

99

Tabl

e 1.

Resu

lts fr

om th

e A

MS

anal

ysis

from

the

diff

eren

t sec

tions

of

the

Rom

ania

n Ca

rpat

hian

fore

deep

. All

the

sect

ions

are

late

Mio

cene

-Plio

cene

in a

ge. N

= n

umbe

r of s

peci

men

s; A

z, D

ip =

mea

n az

imut

h an

d di

p of

axe

s; dA

z, d

dip

= e

rror

s on

the

mea

n k m

ax;

L =

mag

netic

line

atio

n (k

max

/kin

t); F

= m

agne

tic f

olia

tion

(kin

t/k m

in).

100

(Matenco & Bertotti 2000) and NW-SE dextral in the vicinity of the Intramoesian fault(e.g., Zweigel et al. 1998), accommodating a short late collisional ESE -ward movement ofintervening sectors.

Most of the deformation linked with the “Wallachian” (Sandulescu 1988) Pliocene-Quaternary phase of deformation is focused in the SE bending area where N-S to NNE-SSW oriented compression is observed in all paleostress studies (Zweigel et al. 1998;Hyppolite & Sandulescu 1996; Morley 1996, interpreted by these authors to be the resultof a ~N-S oriented contractional event. However, no mesoscale to regional size structurescan be directly correlated to this event of deformation, in all cases high-angle reversefaults oriented WNW-ESE characterize the rather limited area where these structureshave been reported, such as the Breaza anticline or high-angle salt-diapiric structures (e.g.,Stefanescu et al. 2001). The features obvious at regional scale are the N-S oriented FocsaniQuaternary syncline (Lazarescu & Popescu 1986; Matenco et al. 2006) and the high-angle,basement-involved reverse faulting beneath the thin-skinned belt (Bocin et al. 2005). LatestGPS measurement indicate extensional features in relation with the stable Eurasia (vander Hoeven et al. 2005).

3. Anisotropy of the magnetic susceptibility (AMS)Analysis of the AMS can be used to establish the sedimentary and tectonic history

in weakly deformed sediments, because it may indicate the relationship with the regionalstress field of the area (Tarling & Hrouda 1993). In undeformed sedimentary rocks, themagnetic susceptibility is characterized by oblate ellipsoids, with a foliation coinciding withthe bedding plane (i.e. the minimum axes of AMS, kmin, are perpendicular to the beddingplane) and a random orientation of the lineation (i.e. direction of the maximum axes ofAMS, kmax). Upon deformation, the lineation quickly aligns along the direction of maximumextension or, equivalently, perpendicular to maximum compression, i.e. clustering of kmaxin the direction of maximum extension or, equivalently, perpendicular to maximumcompression. The kmin is still perpendicular to the bedding plane. In this study, we measuredthe AMS from 630 samples from 16 different sections, on a Kappabridge KLY-3. Errorellipses of the susceptibility axes are according to (Jelinek 1978) and are given for kmax inTable 1.

4. Materials and sectionsWe used the same paleomagnetic samples as those used for studies by Vasiliev et

al. 2004, 2005, which provided a reliable chronostratigraphy for the Mio-Pliocene interval.Therefore, a good dating of the sedimentary succession was available. Additionally, wealso used extra sites in order to increase the precision of our study (Snel et al. 2006).

The Miocene to Pliocene sedimentary sequences sampled in the Carpathiansforedeep consists of alternations of coarse-grained rocks (sandstones andmicroconglomerates) and fine-grained rocks (siltstones and shales) (Vasiliev et al. 2004;Vasiliev et al. 2005). They have been deposited in a lacustrine to deltaic environment. Themajority of the samples have been taken from riverbeds where the rock surfaces werefreshly cleaned by the stream. The sequences sampled in the eastern Carpathians foredeepare coarser-grained than those from the southern Carpathians Foredeep. Also, they arewell-cemented and display a cyclic pattern which is constant for several kilometers, as canbe observed in the various river incisions that cut through the tilted strata. In the southernCarpathians, the cyclic pattern is not visible at the outcrop scale, which show clear repetitive

101

102

Figure 3. Equal are projections (bedding tilt corrected) of kmax (triangles) and kmin (circles) of theellipsoid of the AMS for individual samples, with the calculation of the mean ellipsoid for eachsection. In the plots are represented also the kint mean values (squares). The divergent arrowsindicate the mean lineation direction per section (calculated from kmax). The convergent arrowsindicate kint. The green arrows represent measurements from the upper Meotian to Romaniandeposits; the red ones are those from the Sarmatian to lower Meotian. ECF, SECF and SCFindicate, sections from the eastern, southeastern and southern Carpathians foredeep, respectively.

103

changes in lithology, from sandstone (rarely micro-conglomerates) to finer–grained rocks.In some sections (e.g. Lupoaia), and mostly in the Dacian deposits, 5 m thick lignite layerswere observed (van Vugt et al. 2001), which are reduced to tens of centimeters in theother valleys like Rîmnicu Sarat, Bãdislava, Topolog, etc.Magnetostratigraphic studies of the eastern Carpathians foredeep cover more than 5Myr, starting in the Upper Sarmatian (~8 Ma) and ending in the Upper Romanian (~2.5Ma). The well-dated polarity pattern served as a tool for calculation of accumulationrates. They indicate high accumulation rates with a sudden increase from 0.6 to 1.55 m/kyr during chron C3r, around 6 Ma (Vasiliev et al. 2004). The increase in accumulationrate roughly coincides with a change in magnetic carriers from iron oxides towards ironsulphide. The magnetostratigraphic record of the southern Carpathians foredeep covers a2.5 Myr time span, from the upper Meotian (7 Ma) to lower Romanian (~4 Ma) stages(Vasiliev et al. 2005). In these sequences, a similar change in magnetic carrier was observedat the same time during C3r.

In addition, we also used samples from Sarmatian-Meotian deposits, which arestill under investigation for magnetostratigraphic results (Milcov, Mãdulari, Cerna andVaideeni sections). Slãnicul de Buzãu, Valea Vacii, Bizdidel, Lupoaia and Ilovaþ were alsocollected along the river valleys, cover each much shorter stratigraphic intervals, andrange from Upper Meotian to Lower Romanian (Snel et al. 2006).

5. AMS ResultsThe AMS results from the Carpathians foredeep generally show prolate and oblate

ellipsoids, the kmin axes being close to the pole of the bedding plane (Figs. 3, 4 and Table 1).The setions from the eastern Carpathians Foredeep (Putna, Milcov, Rîmnicu Sãrat) (Fig. 3a, c, d, e, f) and south-eastern Carpathians Foredeep (Bizdidel Slãnicul de Buzãu, ValeaVacii) (Fig. 3 g, h, m), show significant clustering of the kmax axes. This indicates thatdeformation has been acting on the rock, resulting in the clustering of the kmax in thedirection of maximum extension or, equivalently perpendicular to the maximumcompression. The Late Miocene-Pliocene sediments reveal a roughly N-S alignment ofthe kmax axes, implying N-S extension or E-W compression (Fig. 3 and 5). To the south andeastern, the kmax directions are roughly NNW-SSE oriented. They are parallel to the mainMiocene trusting direction.

The majority of the sections from the southern Carpathians Foredeep (Topolog,Bãdislava, Mãdulari, Cerna, Vaideeni, Bengesti, Lupoaia) (Fig. 3i, j, k, l, m, n, o) arecharacterized by oblate ellipsoids (Fig. 4), with the foliation coinciding with the beddingplane. In these cases, the magnetic fabric is mainly depositional or related to the compactionalloading; the kmin is perpendicular to the bedding plane and the kmax and kint are scattered inthe foliation or bedding plane itself. Most Late Miocene-Pliocene sedimentary rocks indicatean approximately E-W clustering of the kmax axes. The exception is Ilovaþ (Fig. 3p) wherethe direction is oriented NNE-SSW, but however, it respects the general parallel to theMiocene thrust direction of kmax axes observed in the southern Carpathians foredeep.

6. DiscussionsThe distribution of the AMS patterns largely reflects directly the present-day

overall contact between the upper Carpathians plate and the lower platforms situated inthe foreland (Fig. 5). The mean orientation of the AMS lineation directions are orientedNNW-SSE in the eastern Carpathians (Putna site), NNE-SSW somewhat south in Milcov.

104

Rîmnicu Sãrat and Slãnicul de Buzãu have the mean kmax oriented NE-SW. Valea Vacii siteshow a kmax oriented already ESE-WWN. In the southern Carpathians the sites record amean kmax oriented E-W with the exception of Ilovaþ where the kmax is parallel with theorogen and oriented NNE-SSW. The kmax from the eastern Carpathians is basically alignedto the curvature of the South-Eastern Carpathians and parallel to the Quaternary structureof the Focsani basin; in the southern Carpathians the kmax is oriented E-W parallel to theBreaza anticline (Fig. 5).

Our AMS data are generally consistent with the other types of observations andprovides independent support for the inferred pattern of strain (Fig. 5). The paleostresstensors calculated from the faults observed in the field (Morley 1996; Hyppolite &Sandulescu 1996) (Fig. 2) also suggest arc-perpendicular (radial) compression for the regionbetween the orogenic arc and the outer-arc in the, field data imply a prevalence of arc-normal compression radial to the Carpathians arc.

Figure 4. Lineation versus Foliation (Flinn diagram), showing a general prolate to oblate shapefor the AMS measurements in the Eastern Carpathians (solid dots) while in the southern Carpathiansthe ellipsoids shape is strongly oblate (open dots).

Figure 5. (next page) On the same tectono-structural map of the Carpathians as in the figure 3are plotted the AMS data after the bedding plane correction. Shaded segments in AMS dataindicate the errors on the mean kmax in the AMS analysis with solid line as mean lineation directionper area. The darker grey indicates the sites not younger than lowermost Meotian, the lighter greyindicates the sites from Meotian to Romanian rocks. The black arrows indicate the measured GPSvectors from the Eastern Carpathians after (Hoeven van de et al. 2005). The arrow from the legendrepresents a movement of 5 mm/year (±1.5mm/year).

105

GP

S v

ecto

rs d

irect

ions Ilo

vat

p 1-2

Lupo

aia

dc2-

rm1

Ben

gest

i

dc1

Cer

na

sm1-

2

Mad

ular

i

me 2

Vaid

eeni

sm1

Bad

isla

va

p 2-3-

dc1

Topo

log

p 2-3-

rm1

Vale

a Va

cii

m2-

p 1-2

Biz

dide

l

m2-

p 1-2

Sla

nic dc

2-rm

1

Rim

nic

old sm

3-m

1R

imni

c m2-

rm

Milc

ovsm

3-m

1

Put

na

old

sm3-

m1

Put

na

m2-

p

24 E

26 E

22 E

44 N

46 N

-100

0

-500

-150

0

Rom

ania

n&Q

uate

rnar

y

Meo

tian&

Pon

tian

Sar

mat

ian

Mid

dle

Mio

cene

S

arm

atia

n

Fore

deep

(T

he s

tudy

are

a)

LEG

EN

DO

uter

Dac

ides

Sub

carp

athi

ans

Nap

pes

Mar

gina

l Dac

ides

Mid

dle

Dac

ides

Mar

gina

l Fol

ds N

appe

s

Con

volu

te F

lish

Nap

pes

Aud

ia&T

arca

u N

appe

s

Neo

gene

Vol

cani

cs

thru

st o

r re

vers

ed fa

ult

norm

al fa

ult

strik

e-sl

ip fa

ult

Miocene deformation

Pre-Miocene deformation

GP

S v

ecto

rs

dire

ctio

ns

In addition to the general NNW-SSE orientation of the AMS lineation, localdeformation features were observed. The results from Cerna for instance probably, reflectsthe large scale dextral offsets taking place at the end of the Sarmatian, truncating theoverall uplift of Ocnele Mari – Govora anticline. During this phase of deformation, thewestern end of this antiform records dextral transpressional offsets in the order of 4-5km associated with secondary sinistral conjugate faults (Fig. 5), as revealed by detailed fieldmapping correlated with depth studies (Rabagia et al. 2006). This has led to a local re-orientation of the late Sarmatian stress field in the vicinity of this major transpressionalzone. This situation is similar in Mãdulari, where the deformation is associated with south-vergent thrusting of the core of the antiform, as recorded by the Vaideeni AMS site.

A comparable situation is recorded in the Bizdidel section, where large-scale dextraloffsets have been recorded along the Intramoesian fault during the Quaternarydeformations. A local re-orientation of the stress field along an associated NW-SE orienteddextral fault is observed in a local transpressional structure. These dextral transpressionaldeformations take place along a corridor between the Intramoesian fault and the Breazaanticline (Fig. 5) and have led to similar circumstances for the Valea Vacii sections, whoseAMS results indicate an overall WNW-ESE direction.

Slãnicul de Buzãu section marks the transition between the dextral shearing corridorand the overall N-S structure of the Focsani basin. AMS results taken from this localitynorthwards, closely mimic the Quaternary folding axis of the Focsani basin. In placeswhere this axis remains parallel with the overall nappe structure of the eastern Carpathians,measurements in the pre- and syn-collisional sediments have the same orientation with theones measured in post-collisional sediments, such as the Rîmnicu Sarat station. This meansthat the contraction direction leading to the late Sarmatian thrusting of the Subcarpathiannappe was the same as the one recorded by the Quaternary folding.

The orientation of the axis of Quaternary folding generally diverge from theoverall nappe structure of the eastern Carpathians starting from the Putna valley northwards.

Figure 6. Topography of the Pannonian–Carpathian system andpresent-day maximum horizontal stress trajectories (white lines) (afterBada et al. 2001). The (+) and (-) signs mark areas of Quaternary upliftand subsidence, respectively. The dashed line represents the detailedstress trajectory inferred from the AMS data.

106

This has been demonstrated by field and geomorphological studies (Necea et al. 2005) andin surface geology it is visible through the overall Lower Quaternary gravels discordantlycovering sediments as old as the late Miocene over an angular unconformity (Fielitz &Seghedi 2005). This divergence is demonstrated to be the result of a different direction ofshortening, as the AMS results indicate a NNW-SSE direction for late Sarmatian-lowerMeotian sediments, parallel with the nappe structure, and a NW-SE direction of thesubsequent (upper Meotian-Pontian) deposits, parallel with the Quaternary folding axis.

7. ConclusionsAnalysis of the AMS along the eastern and southern Carpathians contact between

the orogenic nappe pile, the lower plate and the overlying sediments indicate that theobserved axes of maximum anisotropy are generally aligned parallel to the Carpathiansorogenic arc.

The frontal shortening in the eastern Carpathians and mainly dextral transpressionalmovements which took place in the southhern Carpathians at the contact with the lowerMoesian plate during the late Miocene have imposed a highly bended plates’ contact. Thisgeometry strictly controls the distribution of the post-collisional stress field in the SouthCarpathians.

Quaternary folding in the eastern Carpathians is characterized by the coexistenceof apparently contrasting styles of deformation and associated vertical movements in arelatively restricted area of SE Carpathians (Matenco et al. 2006; Leever et al. 2006).High-angle reverse faults truncate the lower plate basement and the overlying thin-skinnedunits and are coeval with widespread normal faulting in the distal parts of the forelandlacking a coherent direction of extension (Figs. 1 and 5). This is contemporaneous withlarge sinistral and dextral strike-slip movements along the northern (Trotus) and near thesouthern (Intramoesian) boundaries of the system, respectively, as well as with out-of-sequence oblique “Wallachian” thrusting in the south. Vertical movements involve <5kmuplift of the external nappes and <2km subsidence in the foreland (Sanders et al. 1999;Tarapoanca 2003). All these apparently contradictory types of deformation involve a totalamount of <5km WNW-ESE oriented shortening (Leever et al. 2006). This Quaternary-age deformation is associated with the presence of unusual high-velocity asymmetric mantlebodies (Vrancea slab) and intense seismicity, both apparently occurring in the “wrong”place of the chain (Martin et al. 2006).

In particular, Wallachian deformations, strangely restricted to a narrow dextralshearing corridor between the Intramoesian fault and the high angle, oblique thrusts ofthe Breaza anticline, represent just a local reorientation of the measured paleostress fieldin this corridor and does not reflect a regional deformation event. Therefore, extrapolationof paleostress measurements at regional scale, particularly when at odds with an unusuallyhighly bended orogenic structure must be considered carefully.

Distribution of the present-day stress field in the Pannonian-Carpathians region(Bada et al. 1999; Bada et al. 2001) reflects the pre-imposed plates boundaries contactestablished during the late Miocene collision (Fig. 6). Whether or not this geometry is theresult of Quaternary far-field stresses transmitted by the Adriatic push into the alreadyintraplate type, collision-locked Pannonian-Carpathians system is still a matter to be furtherstudied (Pinter et al. 2005). However, distributions of the post-collisional stresses in theeastern and southern Carpathians apparently do reflect this state of stress. The memoryinherited by the system from the collision times, particularly in terms of deep mantle

107

108

processes (Wortel & Spakman 2000; Cloetingh et al. 2004) is driving only the amplitudesof vertical movements and their evolution through time.

AcknowledgementsThis work was carried out in the frame of activities sponsored by Netherlands

Research Centre for Integrated Solid Earth Sciences (ISES). We also thank to Erik Sneland Mirte Cofino for measuring a part of the AMS data.